Instructions to use mlboydaisuke/Timestral-209M-LiteRT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LiteRT
How to use mlboydaisuke/Timestral-209M-LiteRT with LiteRT:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Timestral-209M on LiteRT: float16 graphs for the GPU and the NPU, int8 graphs for the CPU, host code in Python and Kotlin
11f5cc3 verified Download forms.json from mlboydaisuke/Timestral-209M-LiteRT: direct link, hf CLI and curl.
- Browser
- Download file 2.07 kB
-
https://huggingface.co/mlboydaisuke/Timestral-209M-LiteRT/resolve/main/forms.json
- Command line
-
hf download hf://mlboydaisuke/Timestral-209M-LiteRT/forms.json
-
curl -L -o forms.json https://huggingface.co/mlboydaisuke/Timestral-209M-LiteRT/resolve/main/forms.json
2.07 kB
| { | |
| "model": "Timestral-209M", | |
| "host_tables": { | |
| "forecast_query": "forecast_query.f32", | |
| "rope_cos": "rope_cos.f32", | |
| "rope_sin": "rope_sin.f32", | |
| "rope_first_position": -1023 | |
| }, | |
| "graphs": [ | |
| { | |
| "name": "c2048-h256", | |
| "file": "timestral-209m-c2048-h256.tflite", | |
| "context_patches": 64, | |
| "horizon_patches": 8, | |
| "rows": 2, | |
| "mask": -3e4, | |
| "backends": ["gpu", "npu"] | |
| }, | |
| { | |
| "name": "c2048-h1024", | |
| "file": "timestral-209m-c2048-h1024.tflite", | |
| "context_patches": 64, | |
| "horizon_patches": 32, | |
| "rows": 2, | |
| "mask": -3e4, | |
| "backends": ["gpu", "npu"] | |
| }, | |
| { | |
| "name": "c8192-h256", | |
| "file": "timestral-209m-c8192-h256.tflite", | |
| "context_patches": 256, | |
| "horizon_patches": 8, | |
| "rows": 2, | |
| "mask": -3e4, | |
| "backends": ["gpu"] | |
| }, | |
| { | |
| "name": "c8192-h1024", | |
| "file": "timestral-209m-c8192-h1024.tflite", | |
| "context_patches": 256, | |
| "horizon_patches": 32, | |
| "rows": 2, | |
| "mask": -3e4, | |
| "backends": ["gpu"] | |
| }, | |
| { | |
| "name": "c2048-h256-int8", | |
| "file": "timestral-209m-c2048-h256-int8.tflite", | |
| "context_patches": 64, | |
| "horizon_patches": 8, | |
| "rows": 2, | |
| "mask": -3e4, | |
| "backends": ["cpu"] | |
| }, | |
| { | |
| "name": "c2048-h1024-int8", | |
| "file": "timestral-209m-c2048-h1024-int8.tflite", | |
| "context_patches": 64, | |
| "horizon_patches": 32, | |
| "rows": 2, | |
| "mask": -3e4, | |
| "backends": ["cpu"] | |
| }, | |
| { | |
| "name": "c8192-h256-int8", | |
| "file": "timestral-209m-c8192-h256-int8.tflite", | |
| "context_patches": 256, | |
| "horizon_patches": 8, | |
| "rows": 2, | |
| "mask": -3e4, | |
| "backends": ["cpu"] | |
| }, | |
| { | |
| "name": "c8192-h1024-int8", | |
| "file": "timestral-209m-c8192-h1024-int8.tflite", | |
| "context_patches": 256, | |
| "horizon_patches": 32, | |
| "rows": 2, | |
| "mask": -3e4, | |
| "backends": ["cpu"] | |
| } | |
| ], | |
| "backends": { | |
| "cpu": {"threads": 4}, | |
| "gpu": {"precision": "fp16"}, | |
| "npu": {} | |
| } | |
| } | |